The general focus of this research continues to be the nature of the antigenic determinants on human MHC class molecules that are recognized by cytotoxic T lymphocytes (CTL). Studies conducted during the previous funding period led to the generation and characterization of a panel of mutant HLA-A2.1 molecules, as well as transgenic mice expressing either HLA-A2.1 or HLA-B7. In addition, evidence was developed that implicated peptides derived from the processing of normal cellular proteins ("endogenous" peptides) in the formation of class I associated epitopes recognized by the majority of alloreactive and xenoreactive CTL. In the current application, we will take advantage of a powerful new technique, tandem mass spectrometry, in order to provide detailed information on the complexity and structure of peptides that are produced by the cell and become associated with individual class I MHC molecules. Attempts will be made to establish peptide sequence motifs that determine the ability to bind to HLA-A2.1 or HLA-B7 molecules. The influence of well-defined changes in the structure of the HLA-A2.1 peptide binding site on the specificity of peptide binding will also be assessed. We will also examine tissue-specific variations in the spectrum of peptides associated with HLA-A2.1 and HLA-B7. These studies will be reinforced an analysis of which of these peptides are relevant in the formation of epitopes for alloreactive and xenoreactive CTL, and whether there are CTL that recognize epitopes on class I MHC molecules that are independent of the presence or structure of bound peptides. The second half of this proposal will concentrate on the identification of class I associated peptides and epitopes that derive from viral infection of cells. Again, through the use of the mass spectrometer, we will attempt to identify new endogenously produced viral peptides that are associated with HLA-A2.1 and HLA-B7, and establish their ability to be recognized by CTL. Finally, we will extend our previous results using transgenic mice in order to establish the utility of these mice for thc identification of human class I associated epitopes, and assess their immunogenicity for human T cells. Overall, these studies will approach the nature of peptide-dependent epitopes through an analysis of endogenously produced peptides themselves, rather than through the use of deletion mutants and synthetic peptides. Successful completion of this work should increase our understanding of the origin, structure, and recognition of endogenous peptides bound to human class I molecules, and will provide a basis for further studies on their importance in tissue graft rejection, autoimmunity, and immune responses to virally-infected and tumor cells.